Nanohistory's model reserves the word 'event' for its named graphs which represent historical interactions or assertions of historical activity. They're unnamed by design, allowing users to create large multi-dimensional networks that represent the past. That said, it presents a problem when scholars or users want to label or name historical phenomena, or assert that events belong to a particular happening discussed in documents and evident in the historical record. We've opted for a sub-type of term: 'episodes'. 'Episode' is more flexible than 'event' or 'historical phenomenon', and 'period'. It not only invokes the idea that history is 'episodic', but that the recounting and representation of the past is constructed, narrative, and ultimately something that can be revisited or replayed. And unlike 'period' or 'event', it is indeterminate - it doesn't imply a particular start or stop, and unlike 'period', it doesn't have the shadow of a certain temporality to it.

Because naming historical phenomena is critical to historical scholarship, we've made Episodes into a pseudo-entity type. It's technically a term, but will act like a distinct kind of entity in Nanohistory. You can search for them specifically and they have their own colour. They can be used in tools like other nodes, in particular Webs, Radar, and Narrate, allowing users to capture all of the events and data they include in queries or datasets.

Episodes highlight the constructed nature of history. They are meant to capture how the naming of historical phenomena is detached or removed from the past itself, and how such names can take on their own kind of history. In this way they facilitate Nanohistory's historiographical discourse analysis by acting as topics or terms which can appear in studies and historical prose. Users can track who mentions a particular phenomenon, and who associates what kinds of events with it. This is a substantially distinct approach to other event models, where the event name is required, often along with an event type. 'Episode' allows both the content and descriptions of historical phenomena to remain contestable, flexible, and ultimately reflective of what historical scholarship actually thinks names for events are: labels, nothing more.